Fibroblasts and macrophages constitute major cell populations of the solid tumor stroma, exert pro-tumorigenic functions, and hold considerable potential as therapeutic targets (20). Thus, we aimed to develop 3D tumorigenic spheroids using NSCLC cells, fibroblasts, and monocytes with a potential interest to mimic the natural counterpart. To understand the role of tumor microenvironment in cellular heterogeneity and plasticity, we have also made non-tumorigenic spheroids combining lung epithelial cells with fibroblasts and monocytes in order to compare with tumorigenic spheroids. Briefly, NSCLC cell lines (A549 and NCI-H460 lung adenocarcinoma) or lung epithelial cell line BEAS-2B were cultured in combination with MRC-5 lung embryonic fibroblasts, and THP-1 monocytes in the DMEM media supplemented with 10% FBS and hence named as A549+MRC-5+THP-1 (AMT spheroids) and NCI-H460+MRC-5+THP-1 (HMT spheroids), and BEAS-2B+MRC-5+THP-1 (BMT spheroids) ( Figures 1A–C ). We provided diverse cell counts for spheroid formation across literature and started with three different cell counts per spheroid, i.e., 5000, 8000, and 10,000. To define each cell type’s ratio in the spheroid, we used different ratios of tumor cells: fibroblasts: monocytes, i.e., 1:1:1, 2:2:1, 4:2:1, and 5:2:1 with increasing concentrations of tumor cells and maintenance of CAF density near 40% as reported in the previous literature (21). After optimizing the cell count (10,000 cells per drop), ratio (tumor/normal cells: fibroblasts: monocytes = 5:4:1), and method (hanging drop) ( Figures 1D–F and Figure S1 ), the spheroids were allowed to grow undisturbed for d4, d7, and d10. Developed spheroids were used for the physicochemical and functional characterization by immunofluorescence analysis, whole-mount analysis, and single-cell dissociation for gene expression analysis ( Figure 1G ).

Multicellular 3D spheroid (AMT, HMT, and BMT) diameters were measured after days 4, 7, and 10. We observed an increasing trend in spheroid size and diameters with increasing days ( Figures 2A, B ). BMT and HMT spheroids were 355 ± 3.21 and 347 ± 2.5 μm in diameter on day 4, 425 ± 5 and 405 ± 3.21 μm on day 7, and 514 ± 12.16 and 551 ± 2.8 μm on day 10. AMT spheroids were initially of smaller size as compared to BMT and HMT spheroids, with a diameter of 263 ± 3.6 μm on day 4, 289 ± 3.6 μm on day 7, and 489 ± 2.5 μm on day 10 ( Figures 2A, B ). Moreover, multicellular AMT, HMT, and BMT spheroids formed were of uniform size and diameter, as displayed by microscopic images ( Figure S2A ). The morphology of multicellular spheroids largely depends on cell types and culture methods. Based on cell compactness, spheroid morphology can be categorized as loose aggregates, tight aggregates, or compact spheroids ( Figure 2C ). Loose aggregates have low cell-cell or cell-matrix interactions and are easy to disintegrate, whereas cells are tightly bound to each other in rigid spheroids. The cell adhesion proteins, such as E-cadherin and fibronectin, play a critical role in cell spheroid formation (18). We noticed a marked induction of E-cadherin and fibronectin gene expression with an increase in the number of days in spheroid formation ( Figures 2D, E ). Spheroids were also analysed for N-cadherin expression to test whether this spheroid formation is associated with epithelial-mesenchymal transition (EMT). We found a notable downregulation of N-cadherin expression in AMT and HMT spheroids as compared to BMT spheroids on day 7 and day 10 ( Figure S2B ) suggesting spheroid rigidity. Next, to check the cellular viability within tumor spheroids, calcein-AM/PI staining was performed ( Figures 2F, G ), suggesting cell growth and proliferation of these spheroids over time. We have also determined the cell viability by measuring cytosolic acid phosphatases (APH) activity ( Figure 2H ). However, to confirm cellular proliferation potential within these spheroids, we measured the MKi67 cell proliferation marker gene and Ki-67 protein expression by RT-qPCR and immunostaining, respectively. MKi67 is a nuclear protein that is present during all active phases of the cell cycle (G1, S, G2, and M) but is absent in resting cells (G0) (22). As shown in Figure 2I , a significant increase in MKi67 gene expression was noticed with an increase in days of all three spheroids. However, tumorigenic 3D spheroids showed a much higher cell proliferation rate than non-tumorigenic spheroids. It was also evident from immunofluorescence staining of MKi-67 levels in day 10 tumorigenic spheroids compared to non-tumorigenic spheroids ( Figure 2J ).Moreover, we have sorted fibroblast and macrophage populations from tumorigenic spheroids ( Figures S3A–D ) and analyzed MKi-67 proliferation marker in them. An increasing trend of MKi-67 gene expression was observed in fibroblasts at day 10 whereas induction was noticed in isolated macrophages from day 7 onwards in tumorigenic spheroids ( Figures S3E, F ).

Hypoxic zone and necrotic areas are major characteristics of solid tumors (23). Cancer and stromal cells activate survival signalling pathways in response to rapid oxygen depletion and avascularity (24). Hypoxia in the tumor microenvironment induces stable expression of hypoxia-inducible factors (HIF), which promotes angiogenesis via VEGF induction (23, 25), stimulates pluripotency by increased expression of pluripotent markers (OCT4, SOX2, NANOG) (25, 26), and increases acidosis through enhanced expression of carbonic anhydrase IX (CA-IX) (27). We found a significant increase in the expression of hypoxia-related genes such as HIF-1α, BNIP3, and CA-IX with an increase in the number of days of tumor spheroids development as compared to non-tumorigenic spheroids ( Figures 3A–C ). This was also evident in isolated cell population of fibroblasts and macrophages from tumorigenic spheroids ( Figures S3G–J ). The tumorigenic 3D spheroids exhibited a prominent hypoxic zone, as indicated by the increased Image-iT Green hypoxia staining ( Figure 3D ). Moreover, a significant enhancement of induced pluripotent stem cell (iPSC) markers (OCT4, SOX2, and NANOG) gene expression was observed in tumorigenic 3D spheroids with increasing the number of days as compared to non-tumorigenic spheroids ( Figures 3E–G ) indicating manifestation of pluripotency stimulation in tumorigenic 3D spheroids. The gene expression pattern of these iPSC markers in the isolated fibroblasts and macrophages are not in consistent with the data of tumorigenic spheroids ( Figures S3K–P ) indicating non-stemness nature of the isolated fibroblasts and macrophage cell populations. Further, tumor spheroid-derived cells more aggressively form the colony than the cancer cells in 2D culture ( Figures 3H–J ). These results suggested the development of pathophysiologically relevant TME features within the tumorigenic 3D spheroids.

Fibroblasts and macrophages are the major stromal and immune cell populations, respectively, present within the TME of lung cancer. Fibroblasts are usually found in an inactive form in normal tissues and regulate proper tissue architecture by controlling ECM composition (28). However, when in direct or indirect contact (through secretory proteins) with tumor cells, fibroblasts get activated to multiple subtypes of cancer-associated fibroblasts (CAFs) expressing a unique repertoire of different genes, including collagens and elastins (28, 29). We have isolated fibroblasts from individual 3D spheroids using anti-fibroblast magnetic microbeads ( Figure 4A ) and examined the gene expression pattern of CAF-specific markers. A significant induction of CAF-specific markers gene expression such as α-SMA ( Figure 4B ), FSP ( Figure 4C ), and PDGF-β ( Figure 4D ) was observed in the isolated fibroblasts obtained from day 7 and day 10 of AMT and HMT spheroids as compared to BMT spheroids. Moreover, AMT and HMT spheroids also exhibited an increased level of α-SMA as compared to BMT spheroids ( Figure 4E ). CAF activation in tumorigenic 3D spheroid was further validated with human lung tissue biopsies obtained from non-cancerous and lung adenocarcinoma patients. Increased expression of α-SMA ( Figure S4A ), and FSP ( Figure S4B ) were evident in lung adenocarcinoma samples as compared to non-cancerous tissue samples.

Like CAFs, macrophages residing within the TME adopt a characteristic feature of the CD163+CD206+ tumor-associated macrophage (TAM) phenotype (2, 17,). To assess the model’s capability of recapitulating the immune contexture of TME, such as differentiation of monocytic THP-1 cells into myeloid M2-like TAM phenotype, we have isolated the monocytes from the spheroids Figure 4F and found a striking induction of CD163 (scavenger receptor for the hemoglobin-haptoglobin complex) and CD206 expression ( Figures 4G–I ) in the isolated macrophages obtained from day 7 and day 10 of AMT and HMT spheroids as compared to BMT spheroids. Furthermore, AMT and HMT spheroids also exhibited an increased level of CD163 as compared to BMT spheroids ( Figure 4J ). This data suggests that A549/NCI-H460 lung cancer cells in spheroids promote the differentiation of monocytic cells towards M2-like TAM macrophages. This was further validated in human non-cancerous and lung adenocarcinoma biopsies where lung adenocarcinoma samples showed higher expression of CD163 and CD206 ( Figures S4E, F ) as compared to the non-cancerous samples.

In recent years, significant strides have been made to investigate tumor angiogenesis in 3D tumor spheroids. It caused a paradigm shift in our understanding of how TME influences angiogenic signatures. Interestingly, we have found a massive induction of vascular endothelial growth factors (VEGF) and GATA-2 gene expression in tumorigenic spheroids over time compared to non-tumorigenic spheroids ( Figures 5A, B ). Both VEGF and GATA2 are considered pivotal factors governing endothelial cell development (30, 31). We next sought to analyze the VE-cadherin (CD144) positive (endothelial-specific marker) cells in these spheroids. To our surprise, we noticed a significant number of CD144+ cell population in the tumorigenic spheroids. FACS analysis revealed that 5.61% of cells in the AMT spheroids and 10.63% of cells in the HMT spheroids were CD144+ on the day 10 of these spheroid’s development as compared to non-tumorigenic spheroids ( Figures 5C, D ) indicating the appearance of endothelial cell characteristics in these malignant 3D spheroids. To explore the cell type that gained endothelial features, we have sorted out CD144+ cells from these 3D multicellular spheroids on day 14 and checked the expression profile of fibroblast markers (α-SMA, and FSP) and macrophage markers (CD68, CD163, and CD206). Increased abundance of macrophage markers in CD144+ cells as compared to fibroblast markers ( Figure 5E ) indicated that M2 polarized macrophages are the cell source within the tumor spheroid that could be involved in endothelial cell formation. However, to further examine the potency of these M2 polarised macrophages on angiogenic stimulation, we sorted the CD68+ (pan-macrophage marker) cell population from the spheroids and examined the gene expression profile of different endothelial cells markers and angiogenic regulators. We noticed a profound induction of various endothelial cell marker gene expressions such as VE-Cadherin, VEGF, vWF, CD31, and endoglin ( Figure 5F ), along with the enhancement of gene expression for diverse angiogenic transcription factors like Etv2, FLI1, GATA2, ZEB1, Tie2, and ETS ( Figure 5G ). Moreover, the tumorigenic spheroids displayed cell sprouting when day 14 spheroids were placed on the collagen matrix for seven days ( Figure 5H ) with profound induction of endothelial nitric oxide synthase (eNOS) gene expression ( Figure 5I ). All these results indicated that macrophages within the tumor microenvironment may gain endothelial characteristics and can exhibit features of tumor endothelial cells in the later stages of cancer.

To validate the results obtained from 3D spheroids, we have developed a zebrafish tumor xenograft model by intraperitoneal delivery of the mixed cell population using the same number and ratio as described previously in the spheroid development protocol. A mixed cell population (total cell number 10,000) in 5:4:1 (BEAS-2B/A549/NCI-H460: MRC-5: THP-1) ratio was injected into the peritoneum of immuno-compromised adult zebrafish ( Figures 6A, B ), xenografts were isolated on day 10 and day 14 ( Figures 6C–M ), and the cell population was subjected to gene expression analysis of different markers. Similar to 3D spheroids, we noticed increased expression of the MKi67 proliferation marker in day 10 tumorigenic xenografts compared to non-tumorigenic xenografts ( Figure 6D ). Next, we analyzed the expression pattern of hypoxia-specific marker HIF-1α and its regulated genes such as GLUT-1 and CA-IX. Both AMT and HMT tumorigenic xenografts showed enhanced expression of HIF-1α, GLUT-1, and CA-IX than BMT non-tumorigenic xenografts ( Figure 6E–G ) indicating the development of profound hypoxic zones within day 10 of tumorigenic xenograft development. To check the fate of normal fibroblasts within these xenografts, we have examined the expression profile of CAF markers α-SMA, FSP, PDPN, and TGF-β. The AMT and HMT xenografts showed a marked induction of all these CAF markers ( Figures 6H–K ) compared to BMT xenografts. Moreover, we have also checked the plasticity of THP-1 monocytes in xenografts. Similar to the 3D tumorigenic spheroids, AMT/HMT xenografts exhibited profound induction of TAM population as evidenced by the CD206 expression marker ( Figure 6L ). Similar to the results of endothelial-like cell formation in tumor spheroids mentioned above, we analyzed the gene expression of various endothelial ( Figures 6N–R ) and angiogenic transcription factors ( Figures 6S–V ) in the day 14 xenografts. Results indicated the presence of endothelial-like cell characteristics in these tumorigenic xenografts. All these results suggest the pathophysiological state of AMT/HMT xenografts and show alteration of cellular plasticity in the tumor xenografts model.

Please refer to Supplementary Table 1 : Key resources table for details of resources used for this study.

Human non-small cell lung cancer cell lines (A549 and NCI-H460) and human monocytic THP-1 cell lines were procured from the National Centre for Cell Science (NCCS) Pune, India. Cells were cultured in RPMI1640 media supplemented with 10% FBS and 1% Penicillin-Streptomycin at 37°C in a humidified chamber with 5% CO2. MRC-5 human lung fibroblast cells were obtained from the ATCC, USA; and BEAS-2B human bronchial epithelial cells was a generous gift from Prof. Anita K. Verma. MRC-5 and BEAS-2B were cultured in EMEM and LHC-9 media, respectively, supplemented with 10% FBS and 1% Penicillin-Streptomycin solution at 37°C in a humidified chamber with 5% CO2.

The multicellular tumorigenic and non-tumorigenic 3D spheroids were prepared using the hanging drop method. The detailed protocol of the 3D multicellular spheroid development is given in Figure 1 . Briefly, lung epithelial cells (BEAS-2B) or lung adenocarcinoma cells (A549 and NCI-H460) were combined with MRC-5 lung fibroblasts and THP-1 monocytes in a complete growth medium (DMEM, 10% FBS and 1% Penicillin/Streptomycin). Developed spheroids are named BMT (BEAS-2B+MRC5+THP-1), AMT (A549+MRC5+THP-1), and HMT (NCI-H460+MRC5+THP-1). Cell suspension was prepared using BEAS-2B/A549/NCI-H460 cells in combination with MRC-5 and THP-1 in a 5:4:1 ratio, i.e., 5000:4000:1000 cells/ml and 25μl (cell suspension in 5:4:1 ratio and complete DMEM medium)droplets were spotted onto 90mm cell culture dish lid. We have standardised 25μl as maximum volume of cell suspension to prepare a droplet as it will not fall down while inverting the lids. Approximately 50 droplets can be placed on each 90mm dish. The lid was then gently placed on a plate without disturbing the droplets, and 6 ml of autoclaved water was added to the dish’s bottom to keep the cells hydrated. Plates were maintained at 37°C in a humidified incubator with 5% CO₂ for 4 days to allow the spheroid formation. After day 4, spheroids were routinely monitored under the microscope and imaged to examine cell aggregation and proliferation. For a live/dead assay, five spheroids each were collected in an eppendorf tube on day 4, day 7, and day 10, gently washed with PBS and incubated with 1μM Calcein-AM and 2 mg/ml propidium iodide solution for 10 min. On termination of the incubation period, spheroids were gently washed with PBS twice, pipette out on a glass dish using 50μl tips, and imaged using an inverted fluorescent microscope (Leica DMi8, Germany).

Human lung tissue biopsies were collected from the patients with or without non-small cell lung cancer (adenocarcinoma) from the Postgraduate Institute of Medical Education and Research (PGIMER). The study protocol was approved by the Institute Ethics Committee (IEC), PGIMER, Chandigarh. The Declaration of Helsinki protocols were followed and informed written consent was obtained from all the patients. Demographic details of the patients are mentioned in Supplementary Table 2 .

Adult zebrafish (Danio ratio) of both sexes were maintained under a 14 h/10 h day/night photoperiod at 28 ± 2°C in 4L fish tanks containing autoclaved, sterilized water supplemented with penicillin-streptomycin. Water was changed every day and fish were fed twice a day with commercially available feed (MicroMac, Aqua World, India). The Zebrafish lung cancer xenograft model was developed following a previously described method with slight modifications (45). Briefly, caerulomycin (100 mg/Kg body weight) was administered in the intraperitoneal region of zebrafish for immunosuppression, 3 days prior to cell transplantation and fish were maintained in water containing 1% penicillin-streptomycin. On day 3, 10⁴ cells/fish [5 (BEAS-2B/A549/NCI-H460): 4 (MRC-5): 1 (THP-1)] were suspended in PBS and injected into the peritoneal cavity of the zebrafish using a 5μl Hamilton syringe (Hamilton, Nevada USA). Subsequently, fish were maintained in distilled water with 1% penicillin-streptomycin for the next 10 or 14 days allowing for tumor xenograft development. On day 10 and day 14, fish were sacrificed, tumor tissue was harvested and used for downstream experiments. The protocol and procedures were approved by the Institutional Animal Ethics Committee, Kirori Mal College, University of Delhi (Protocol no: DU/KR/IAEC/ZF2021/1).

Acid phosphatase (APH) activity of the spheroids was determined following the method described by Friedrich et al (46), using 4-nitrophenyl phosphate as a substrate. Briefly, spheroids were washed with PBS and placed in a 96-well plate (10 spheroids/well). APH assay buffer (100 µl), containing para-Nitrophenyl phosphate (PNPP, 2 mg/ml), Triton-X (0.1%) in Citrate buffer (0.1M), was added, and the plates were incubated for 90 min at 37°C. On termination of incubation, NaOH (1M, 10 µl) was added to each well, and absorbance was measured at 405 nm on a microplate reader (Multiskan GO Microplate Spectrophotometer, Thermo Fisher Scientific, Finland).

Total RNA was extracted from the spheroids, isolated cells, human tissue, and zebrafish xenograft samples using Trizol reagent according to the manufacturer’s instructions. 100 spheroids per condition were collected in an eppendorf, tube, gently washed with PBS, disintegrated using 200μl of 0.25% trypsin-EDTA solution for 15 min at 37°C, and centrifuged at 1200 rpm for 5 min at 4°C to proceed further for RNA isolation. RNA quality was measured using NanoDrop One/One Microvolume UV-Vis spectrophotometer (Thermo Fisher Scientific, USA) and treated with DNase I. For mRNA expression analysis, cDNA was prepared from 500ng of total RNA using the iScript cDNA synthesis kit, following the manufacturer’s guidelines. Single-cell RNA isolation was performed using a single-cell lysis kit following the manufacturer’s protocol. 100ng of total RNA was reverse transcribed using SuperScript VILO cDNA synthesis kit. PowerUp SYBR^® Green Master Mix qPCR (2X) Universal was used to perform RT-qPCR analysis in a Quant-Studio 5 Real-Time PCR System (Applied Biosystem, USA) to quantify the relative mRNA expression level using gene-specific primers. After the final extension, a melt curve analysis was performed to ensure the specificity of the products. Data were normalized to the expression of the GAPDH reference gene. Primer sequences used for RT-qPCR are listed in Supplementary Table 3 .

Immunostaining was performed on spheroid cryosections using specific antibodies. Briefly, spheroids were washed with PBS, fixed with 4% paraformaldehyde for 15 mins, and embedded in 1% agarose solution (20 µl agarose droplet was used to coat each spheroid. The spheroids can further be made visible in OCT by colouring agarose solution with ponceau stain. Spheroids in agarose droplets were then embedded in -30°C cryofixed, in OCT, and kept at -80°C before further use. OCT-embedded spheroids were cryosectioned (6 microns), fixed with cold-acetone, blocked with 5% Bovine Serum Albumin (BSA), and incubated overnight with specific primary antibodies. Signal was visualized by subsequent incubation with fluorescence-tagged appropriate secondary antibodies (Alexa 594-tagged anti-mouse, 1:1,000 dilution; Alexa 594-tagged anti-rabbit, 1:1,000 dilution) and counter-stained with DAPI. Images were captured by a fluorescence microscope (Leica DMi8, Germany), and analysis was performed using LASX software.

To determine the clonogenicity of multicellular 3D spheroids, day 10 spheroids were disintegrated using 0.25% trypsin-EDTA solution for 15 min, centrifuged at 1200 rpm for 5 min at 4°C, and cell pellets were washed and resuspended with complete DMEM medium. Cell suspension was then seeded at a density of 4000 cells/60mm dish and allowed to grow for the next 14 days. To determine the clonogenicity of BEAS-2B, A549 and NCI-H460 cells in 2D culture, cells (4000 cells/60 mm dish) were seeded and continued to grow for 14 days. On day 14, culture media was removed, and cells were washed with PBS twice and fixed with 4% paraformaldehyde for 15 min. After fixation, cells were stained with 0.5% crystal violet staining solution at room temperature for 20 min, and then plates were rinsed with distilled water, air-dried, and photographed.

Spheroids were disintegrated using 0.25% trypsin-EDTA solution at 37°C for 15 min, centrifuged at 1200 rpm for 5 min at 4°C and cell pellets were washed with chilled PBS. Cell pellets were then re-suspended in cell staining buffer (PBS pH 7.4 + 0.2% FBS+ 0.09% NaNO₃) and blocked with Fcγ blocker (TruStainFcX™, Biolegend). Cells were incubated with PE conjugated VE-Cadherin (anti-human), APC conjugated CD163 (anti-human), APC conjugated α-SMA (anti-human), APC conjugated CD68 (anti-human) antibodies for 1 h in rotatory mixer at 4°C. Labelled cells were washed twice with PBS, re-suspended in cell staining buffer and analyzed in a Flow cytometer (BD Accuri C6+, BD Biosciences, San Jose, CA) using FlowJo™ v10.6.1 software.

For positive selection of single-cell using magnetic beads, approximately 100 BMT, AMT, and HMT spheroids were taken per condition in an eppendorf tube. Spheroids were washed once with PBS and disintegrated using 0.25% trypsin-EDTA solution at 37°C for 15 min, centrifuged at 1000 rpm for 5 min at 4°C, counted, and 10⁷ cells of each spheroid were resuspended in 80μl of a cold buffer containing PBS, 0.5% BSA, and 2mM EDTA. CD11b/CD68, VE-cadherin, and anti-fibroblast magnetic bead separation (Miltenyi Biotech) were done according to the protocol provided by the manufacturer. Briefly, 20μl/10⁷ of the respective antibodies were added per condition to isolate CD11b, VE-cadherin, and α-SMA positive cell population, mixed well, and incubated for 30 mins at room temperature. Cells were washed by adding 1ml of cold buffer, centrifuged at 1000 rpm for 5 mins, the supernatant was aspirated, and the cells were resuspended in 500μl buffer. For magnetic separation, the MACS column was prepared by rinsing with 3ml buffer, the cell suspension was applied to the column, and flow-through containing unlabelled cells was collected. After washing the column 3 times with 3ml buffer, the column was removed from the separator and placed on a collection tube. Magnetically labeled cells were collected by adding 5ml buffer and by firmly pushing the plunger into the column to get the desired cell population. Flow-through was collected to obtain tumor cell population, centrifuged, and used for further analysis. CD68+ cells were isolated using protein G magnetic beads. Protein G magnetic beads were tagged with specific antibodies for 2 hours at 4°C. Disintegrated spheroids were incubated with the tagged magnetic beads for 2 hours at 4°C, centrifuged at 1000 rpm for 5 minutes to remove excess antibodies, resuspended in PBS, and positive cells were isolated by magnetic separation.

All data analyses were performed using GraphPad Prism software (v.8.0; GraphPad Software, Inc., La Jolla, CA, USA). Data represented as mean ± S.D. Students t-test determined statistical significance, and p value indicated significance. Data represent as mean ± S.D. A level of p< 0.05 was considered significant.